2019-01-10 | Mengzhenyu (Zhenyu) Zhang：Creating demand hype with pricing and availability: Analysis and approximation
Traditionally, dynamic pricing models take an "isolated'' view of customer behavior by assuming demand is independent and the firm can price each potential purchase individually. With the rise of "social'' technologies that quickly and intimately connect consumers, this traditional view is being challenged. Consumers within a social media circle can significantly influence each other's preferences. Moreover, social media platforms can simultaneously notify followers of new content, leading to the potential synchronization of purchases and challenging the view that each purchase can be priced individually.
We propose a modeling framework for dynamic pricing that embraces the challenges and opportunities afforded by this emerging "social'' economy. One goal of social media marketing is to quickly create "hype'' within the niche followings of social influencers. Our framework can be used to dynamically "price hype'' by allowing demand to depend on past sales, initial availability, and current inventory. We also allow for periodic pricing policies, where the number of price changes is an input, acknowledging the fast-moving nature of demand in niche markets.
To analyze problems in this framework, we propose a certainty equivalent (CE) heuristic policy that takes as input only the conditional expectation of demand for a single good between each price change. Our CE policy performs well asymptotically: as the potential market size m for a product increases, the percentage loss as compared to the optimal policy converges to zero sublinearly. Our numerical results show that even with relatively small values of m and very few price changes, the CE policy performs impressively. Finally, we look at the joint optimization of pricing and initial inventory and leverage our CE analysis to provide asymptotically optimal recommendations on whether and how much to "underserve'' the market by restricting initial inventory.
(Coauthor:Hyun-Soo Ahn, Christopher Ryan, Joline Unichanco)
Mengzhenyu (Zhenyu) Zhang studies Technology and Operations at the Ross School of Business, University of Michigan. She is currently a fourth year PhD candidate. Her research advisors are Professor Hyun-Soo Ahn and Professor Joline Unichanco.
Her current research mainly focuses on studying operations decisions (pricing, inventory management) by considering the non-stationarity of the market (such as word-of-mouth effect or scarcity effect in product demand). The focus of research is using analytical tools such as optimization and statistical learning to develop policies that are implementable in realistic settings.